Published in The Australian, Higher Education, Wednesday 10 October 2012
Last week, Simon Marginson wondered why the Grattan report Graduate Winners did not reference OECD data (HES, October 3). In hindsight we should have explained why we did not use this source, given how popular OECD comparisons are in Australia, writes Andrew Norton
But it was a deliberate omission. OECD statistics are not always the most appropriate for issues such as those examined in our report.
Graduate Winners asks what Australia gets for its public investment in higher education that it would not get anyway from a more user-pays system. One theory is that public subsidy attracts more students with lower fees, and via more graduates produces additional benefits partly captured by the general public. These include increased tax revenues and non-financial benefits, such as more volunteering or more tolerant attitudes.
Marginson’s article focuses on the non-financial benefits, suggesting that “inconveniently” OECD statistics for other countries find larger effects than Graduate Winners reports.
The OECD publication Education at a Glance includes what percentage of various age cohorts in different countries acquire degrees and information on the fees paid by students at public universities. If we correlate the two data sources, we find that countries with more expensive higher education tend to have more people with degrees. So it calls into question the idea that public spending to lower fees is essential to attracting more people to university.
The risk with public funding is that all governments have limited funds for higher education. They therefore need to ration it. They can ration by limiting how many people receive a subsidy. They can also limit how much each person receives. With a fixed pot of money, the second method helps more people, and other things being equal leads to more graduates. There is a potential paradox: higher education subsidies increase demand but may decrease supply.
While the risk of this paradox exists, it hardly settles the issue. The OECD’s figures show that high attainment is possible in either high or low fee systems. Rather than conclusively demonstrating that one way is better than the other, it shows the contextual nature of higher education funding systems. With Scandinavian tax levels, free education is compatible with high attainment levels. High fees can work in lower-tax countries where families have more disposable income and graduates keep more of their education’s financial benefits.
In the Australian context, a major issue is the effect of higher fees on net private financial benefits, which provide an incentive to study for many (though not all) students. Graduate Winners modelled various student charges, up to average international student fees. Australian earnings profiles and tax rates are essential to this exercise. Inference from OECD surveys just cannot answer the crucial questions.
For the non-financial public benefits of higher education, we started with the list in last year’s base funding review and its accompanying background paper. The list came from an international literature review. That review was a useful starting point, but before we say (as the review did) that Australian taxpayers should pay for these benefits, we should at least check on their presence and scale in Australia.
So Grattan did this research, using data from major social surveys such as HILDA and the ABS General Social Survey. Generally, we found that graduates were more likely than non-graduates with similar backgrounds to engage in various pro-social behaviours. But often the differences are smaller than in the United States, the source of most research in this area. Marginson seems to prefer the American data. The US may show what can be achieved. But it does not reflect contemporary Australian reality, and so should not be used as the basis of Australian public funding.
Marginson presents me and my Grattan colleagues as ideological hacks who selectively chose evidence that supports pre-determined conclusions. But I think anyone who reads Graduate Winners and its accompanying technical reports will realise that together they are a comprehensive analysis of these issues. On many topics, we could and did use multiple datasets to ensure that our results were robust. We have published our methodology, so other social scientists can check our findings.
As with all social science, our empirical findings can be questioned with new data or other statistical methods. But it is notable that to date nobody has done so, other than pointing to OECD surveys which, on examination, do not even clearly support their case. If Professor Marginson wants to disprove the empirical evidence in Graduate Winners, he will have to do a lot more than just selectively summarise Education at a Glance.